Weird Buckaroo Test Lead Physics??? Moderators: adafruit_support_bill, adafruit

[mnopneal]

I'm passing this along as it's maybe a gotcha for the unspecting. It's not exactly cold fusion, but seems to defy the laws of electronics. My Banzai Buckaroo board interfaces to a Micro:bit controller. It measures Pin 1 against 3V to judge soil moisture, and below a threshold activates a 3V circuit to power a pump or solenoid.

HERE is the freaky part: I was thinking I'd make a sturdy test lead. Adafruit suggests 2 stainless steel nails and alligator clips. That felt like high-school science, not production use in a garden bed. SO, I had some 6 wire cable, I twisted three of each together, and tinned the ends, including 3" on the end we stick in the soil. I tested it, and there is infinite resistance between each set.

And it worked badly. The working range on Pin 1 is apparently from 0 to 1023. Pin 1 Disconnected reads 234. Pin 1 connected, but the leads sitting in the air, reads 794!!. With leads crossed it reads 1023. Stuck in soil which is dry it reads 975. Useless.

BUT, connect a separate piece of twin-lead wire to Pin 1 and it reads 357!!???. Wuh? Connect both leads of that twin-lead wire to 3V and Pin 1, and it reads 697???

Cut the twin-lead wire down to 4' long, and it reads 511?? Pull the very same wire apart and it reads 340?? Some bizarre coupling is going on which exceed my high-school electronics.

So, make sure your test leads aren't close to each other.

[adafruit_support_bill]

Pin 1 Disconnected reads 234. Pin 1 connected, but the leads sitting in the air, reads 794!!.

No laws of physics being broken here. The voltage on an unconnected pin is unpredictable. It will report whatever static charge is on the pin, which may be influenced by charges or electromagnetic fields on nearby objects.. Adding a bit of wire with the other end hanging in the air just gives you a better antenna for picking up random stuff.

[mnopneal]

I guess what I am seeing is that the length of the wires, and whether the wires are near each other has a profound effect on whether the readings are usable.

Using 4' wires, stripped 2.5", inserted into soil that needs water:
7" apart Pin1 reads 999
14" .............................1002
23".............................995

Cutting the probe leads down to 1 foot long, and stripped less than an inch:
With Pin 1 lead in the air we read 325
With Pin 1 in the soil 16" from the other lead we read 960.

IF we then water the garden bed it goes to 993.

SO, if we set our threshold to 975, then MAYBE this will work.

IF we make our test leads longer, maybe not.

IF we strip off more copper, maybe not.

IF we have a longer probe, maybe not.

It seems to be on the hairy edge of working.

[adafruit_support_bill]

With the circuit shown, pin 1 in the air is meaningless as explained above. It has no connection to any known voltage so its value is unpredictable.

With both 3.3v and pin 1 in the soil, you have a very high resistance connection between pin 1 and 3.3v, so the value should be at or close to full-scale.

What is missing is a resistor between pin 1 and GND. That would create a variable voltage divider with the soil as the variable part. So the voltage at pin 1 would be a function of the ratio between the fixed resistor value and the soil resistance. 10K is probably a good value to try.

[mnopneal]

Thanks for the reply.

I will certainly try a 10K resistor.

Do you have thoughts about optimal probes for a raised garden bed? More surface area? Maybe a flat steel bracket with the copper lead screwed down?

I'm not looking for perfection; just a fairly reliable indicator of when to pump in some more water.

regards,

Neal

[adafruit_support_bill]

Unprotected steel is going to rust away pretty quickly. There are lots of tutorials that use galvanized nails. But those are probably better suited for something smaller-scale like a potted plant. For the garden, some galvanized garden stakes might be a better choice. You could use a small ground clamp or cable clamp to attach the wire.

[mnopneal]

Thanks for the reply.

I will certainly try a 10K resistor.

Do you have thoughts about optimal probes for a raised garden bed? More surface area? Maybe a flat steel bracket with the copper lead screwed down?

I'm not looking for perfection; just a fairly reliable indicator of when to pump in some more water.

regards,

Neal

[mnopneal]

Using an 11K resistor, just using .5" copper: Pin 1 open reads 3, Stuck into dry soil by the garage door, 107, same soil after watering: 178

Using a 36K resistor, and bolting the leads to a pair of 8" galvanized mending plates: Using same dry soil, Pin 1 open reads 20. In dry soil reads 316, in wet soil reads 353.

SAME resistor and leads, moving up to the garden bed on the deck, in dry planting soil reads 916; after watering reads 976.

So, wildly different results in different soil mixtures. As long as it's consistent we can calibrate to something useful.

regards,

Neal